Artificial intelligence automatic defect coding for cleaning videos

Sewer Robotics has added automatic defect coding using artificial intelligence for videos recorded with the C70 video nozzle in the (free) Sewer Analytics web cloud.

Cleaning and measuring pipes can be an extremely tedious job, but there are ways to streamline the process. From recording high-resolution videos with a cleaning nozzle to training AI-powered software to automatically detect faults, Sewer Robotics has developed effective solutions that make cleaning and measuring large quantities of pipes and reporting their faults easier and more efficient, even without human eyes reviewing the videos.

See how AI can be used to automate defect coding in cleaning videos and how it can help streamline the pipeline asset management process, from data collection and analysis to efficiently prioritizing actions and defining what to take.

C70 Video Nozzle

With the advancement of technology and engineering, new ways to clean pipes with modern equipment have become available. One such device, the C70 Video Nozzle, is a pipe cleaning jet nozzle with integrated front lighting and a battery-powered, self-leveling, high-resolution camera for recording during pipe cleaning.

The C70 video nozzle is made of robust stainless steel and can be compatible with any jetter by screwing onto the hose. The integrated pressure switch activates the lighting and starts recording when pressure is built up, allowing the cleaning worker to concentrate on operating the jetter and cleaning the pipe.

After the pipe is cleaned, the video can be downloaded to instantly visualize the pipe condition and help the operator choose the right cleaning method. And in most cases, if the cleaning goes without anomalies, the video is not reviewed on-site. However, at the end of the day, it is automatically uploaded to the Sewer Analytics online cloud along with the GPS coordinates, time and date of cleaning.

Sewer Analytics C70 video nozzle observation.

AI is added to clean videos

Artificial intelligence, or AI, refers to machines that can make intelligent decisions. These machines are not robots with arms and legs, but rather software programs that make our lives easier by performing tasks that normally require human intelligence, such as understanding natural language, responding to images and sounds, and playing games.

Artificial intelligence can also be used to automatically encode defects in cleaning videos. This is done by first training the AI ​​for pipe defects using a large set of encoded defects. After training, the AI ​​then looks for patterns in the new video that match the patterns it learned from the defective videos. Finally, the AI ​​outputs a list of defects for the cleaning video, along with the required screenshot and defect code according to international standards.

Defects detected by AI

You don't even need to watch the video because the AI ​​will inform you that there are grease deposits and bubbling intrusions in the pipe and classify the pipe segment as high priority #5. Other commonly reported defects include roots, displaced joints, surface damage, cracks, breaks, deposits and penetrating joints.

Canal analysis c70 infiltration and fat

AI for wastewater infrastructure management

As the world becomes more digital, companies are turning to artificial intelligence (AI) for a variety of tasks. AI is even being used in wastewater infrastructure management. By using sensors and data analytics, AI can help identify problems with stormwater, water, and wastewater systems before they become problems. It is expected to revolutionize our industry, saving companies time and money by preventing damage to infrastructure and avoiding business disruption.

When it comes to pipe cleaning and inspection, you can use a video nozzle like the C70 and record a video in HD format while cleaning, which you can download once the cleaning process is complete. In fact, it allows asset managers and cleaning teams to record high definition videos while doing pipe cleaning operations at a minimal cost.

Economical advantages

Video nozzles are devices for simultaneous pipe cleaning and video recording and can be used during regular cleaning or pre-inspection instead of conventional cleaning nozzles on any high-pressure cleaning vehicle and do not require any methodological changes. (Video) data collection is therefore carried out with minimal additional effort and cost compared to regular cleaning, which is always required before inspection.

Since this represents only a small part of the system (estimated at 15 – 30%), additional measures are required and therefore using a video nozzle to capture video data results in a significant cost reduction.

Although a video nozzle cannot completely replace a detailed CCTV crawler inspection, it is much more cost effective and will show you if there are any defects in the pipe. Depending on the reported defect and its priority, the correct actions can be defined and implemented based on the cleaning video. In the few cases where a closer look is desired, a detailed CCTV crawler inspection can also be carried out if desired.

And an HD video nozzle like the Sewer Robotics C70 provides video evidence that pipe sections are completely clean and problem-free, so resources can be deployed where they are needed.

Visit Sewer Robotics for more information and videos.

You might also like

Comments are closed.